Why I declined to clone an app

The brief was to clone an existing Mac disk-cleanup product exactly: the same interface, the same feature set, the same look. Whether it is legal to clone an app like that comes down to what gets copied. The features are the easy part, since the idea of scanning a folder and showing what is taking up space is not something anyone owns. The name, the branding and an interface built to be mistaken for the original are a different matter, and on trademark grounds the exact copy was declined.

What got built instead was DiskScout, an original native Mac app matching the same feature set under its own branding, in a category that already has several players. It scanned a folder tree with live progress and rendered it as a list, a treemap, a sunburst, ranked sizes and an age map. It had a duplicate finder, showed a file’s real size on disk, and staged deletions in a queue that removed nothing until it was explicitly emptied.

That queue was the one design decision worth defending on its own. A disk cleaner’s worst failure is deleting something a person needed, and the fastest interface for the job makes deletion a single click. Staging turns a destructive action into a batch someone reviews before committing, which costs one extra step and removes the kind of mistake that makes people afraid of the tool in the first place.

It never shipped. The app compiled, launched and ran with a working interface, and then the scan pipeline deadlocked under load partway through end-to-end testing. That bug was mid-fix and unverified when the entire project directory was deleted, so there is no build to show and no fix to confirm.

The case study stays on the site anyway, with the deadlock and the deletion described plainly, because neither of the two decisions in it, declining the copy and declining one-click deletion, depended on the app shipping.